Fluent Essays

Summary of the Article attached should include: In 1-2 pages please complete the following  1) Purpose:  a brief statement of the main point or purpose of your study 2) Methods (measurements and sample characteristics): What and who did they measure? What was the main theory used? .. This article uses  multivariate analysis please include that in the methods as well 3) Results: Summarize only the main findings in your own words! 4)Conclusions/Critique - What did you think of the study and its methods? How could it be improved? Full Reference in APA format Many business publications and journals require APA style.   Here are some style guides:  http://www.lib.sfu.ca/help/cite-write/citation-style-guides/apa/apabusiness (Links to an external site.) http://libguides.lib.msu.edu/c.php?g=96008 (Links to an external site.) https://owl.english.purdue.edu/owl/resource/560/01/ (Links to an external site.) Child Adoption in the United States: Historical Trends and the Determinants of Adoption Demand and Supply, 1951-2002 Raquel Bernal, Universidad de Los Andes Luojia Hu, Federal Reserve Bank of Chicago Chiaki Moriguchi, Hitotsubashi University Eva Nagypal, Northwestern University First Version: December 2007 This Version: November 10, 2009∗ JEL Classifications: J13, N32, D19 Abstract Adoption, as an alternative to child bearing, is a widely accepted means of forming a family in many modern societies. In this paper, we first provide a comprehensive overview of the U.S. adoption market and its historical development. We then document trends in adoptions in the U.S. by adoption type using aggregate-level data from 1951 to 2002 and explore possible reasons for the observed historical patterns. Finally, compiling two micro-level datasets covering the period between 1973 and 2002, we estimate individual women’s propensities to adopt and to relinquish a child for adoption. In our empirical analysis, we evaluate alternative hypotheses concerning adoption demand and supply, exploiting both across-time and across-group variations in the data. ∗ Corresponding author: Chiaki Moriguchi, Institute of Economic Research, Hitotsubashi University, 2-1 Naka, Kunitachi, Tokyo 186-8601, Japan; E-mail: [email protected] We thank seminar participants at UC Irvine, UCLA, and Northwestern University for helpful discussions and Camilo Bohórquez, Pamela Giustinelli, Lance Kent, and Tuan Hwee Sng for excellent research assistance. Financial support from the National Science Foundation (grant SES-0721137) and Northwestern University (URGC grant) is gratefully acknowledged. 2 1. Introduction Adoption, as an alternative to child bearing, is a widely accepted means of forming a family in many western societies. According to the 2000 Census, adopted children comprise roughly 2.5\% of all children in the U.S. (U.S. Census Bureau (2000)), while unmet demand for adoption is reported to be far greater (NCFA (2007)). The number of foreign-born children adopted by U.S. citizens, in particular, has more than tripled in the last fifteen years (U.S. Department of Homeland Security (2005)), generating wide press coverage. Although there are no internationally comparable statistics, data indicate that the U.S. likely adopts more children per capita than any other country in the world (Selman (2002)). Who adopts children, and who places children for adoption? How does the market for child adoption function? In spite of rapidly declining demand for biological children in most industrial countries (measured by total fertility rates), what motivates people to adopt children who are not biologically related to them? Is the number of adoptions rising in the U.S.? If so, is the trend driven by economic and demographic factors, or is it more attributable to legal and institutional factors? Despite its quantitative significance and potentially important welfare implications, unlike sociologists and demographers, economists have paid relatively little attention to child adoption due partly to the scarcity of data (Fisher (2003)).1 Child adoption, however, is no less an economic issue than is child bearing in the classical Becker (1981) model. The objective of this paper is threefold. First, we provide an overview of the U.S. adoption market, its institutional features, and its historical development over the last 150 years. Second, we document adoption trends in the U.S. by adoption type from 1951 to 2002 using macro (aggregate-level) data compiled from several sources. To explore possible reasons for the observed historical patterns, we divide the adoption market into three submarkets and examine the demand- and supply-side factors in each segment. Third, we investigate the determinants of adoption demand and supply using two micro (individual-level) datasets, the Survey of Income and Program Participation (SIPP), 1984-2001, and the National Survey of Family Growth (NSFG), 1973-2002. By providing a comprehensive overview and new empirical evidence, we hope to advance the understanding of child adoption, an important yet thus far little understood subject in family economics. 1 Important exceptions are Landes and Posner (1978), Medoff (1993), Gennetian (1999), Hansen and Hansen (2006), and Buckles (2006b). 3 2. Institutional Background Child adoption has a number of features distinct from child bearing that require careful investigation. Unlike in the case of child bearing, individuals on the demand side (i.e., adoptive parents) and on the supply side (i.e., relinquishing parents) do not coincide, giving rise to an “adoption market” in which the two sides are matched by intermediaries (i.e., adoption agencies) using primarily non-price mechanisms. As the adoption market deviates greatly from the standard competitive market, institutional factors, such as adoption laws and cultural norms, may play an important role in determining its outcomes. In addition, adoptive parents may incur monetary, time, and psychic costs that are specific to adoption (e.g., legal costs, long waiting time, great uncertainty over the quality of match between a prospective parent and a child, possible stigma associated with adoption), or they may be motivated by factors other than those driving child bearing, such as humanitarian considerations. Moreover, there are several types of adoption that are imperfect substitutes for one another in the adoption market. Adoptions can be categorized as formal versus informal adoptions, related versus unrelated adoptions, domestic versus inter-country adoptions, and foster care adoption. Formal adoption refers to legally approved adoption that entails permanent transfer of parental rights and duties from biological to adoptive parents. In the U.S., state courts decide whether to grant or deny a petition for adoption. Related adoption includes adoptions by relatives as well as stepparents, while unrelated adoption refers to the rest. Unrelated adoption, especially of young infants, is considered to be a close substitute for child bearing, while it is less likely to be the case for related adoption, especially for stepparent adoption. As we show below, related adoption constitutes roughly 40\% of all adoptions in the U.S. in recent years. Furthermore, adoptions can be divided into domestic and inter-country adoptions. For jurisdictional reasons, most domestic adoption (i.e., adoption of U.S. children by U.S. citizens) takes place within state, and inter-state adoption is relatively rare. Inter-country adoption refers to adoption of foreign- born children by U.S. citizens. Almost all inter-country adoptions are unrelated adoptions. Finally, within domestic adoption, there is foster care adoption. Over the last two decades, the federal government has placed an increasing emphasis on finding adoptive homes for children in the public foster care system, especially for those children with “special needs.” The definition of special needs children varies from state to state, but it typically refers to children who are above a certain age, members of a sibling group to be adopted together, of minority race, or have physical, mental, 4 or emotional disabilities. Foster care adoption can be either related or unrelated adoption. Because the characteristics of adoptable children and adoptive parents differ systematically across adoption types, it is important to distinguish these types as much as data permit. In the U.S., adoptions can be arranged through public child welfare agencies, private agencies, or without involving any agencies (i.e., “independent” adoption). Almost all foster care adoptions are arranged through public agencies. [Need more details.] Data suggest that a majority of independent adoptions are adoptions by stepparents and relatives. [Misleading statement: revise.] By contrast, almost all private agency adoptions and a majority of public agency adoptions are unrelated adoption (Flango and Flango (1995), Table 3). The primary functions of adoption agencies are to evaluate prospective applicants, conduct home studies, arrange suitable placements, and process court applications. A birthmother may relinquish all rights to her child to an agency. All private agencies are licensed and subject to state regulations. Most private agencies are non-profit organizations, but some states permit pro-profit agencies (O’Halloran (2006)). Monetary and time costs of adoption vary substantially by agency and by adoption type. Estimated monetary costs in 2004 range from $0 to $2,500 for adoption of foster care children through public agencies, $5,000 to $40,000 (with the average $10,000-$20,000) for domestic adoption through private agencies, and $7,000 to $30,000 for inter-country adoption through private agencies (CWIG (2004)). Expected waiting time for adopting healthy domestic infants through private agencies is between 2 and 4 years, reflecting high excess demand for these children (NCFA (1989)). Expected waiting time for adopting healthy infants from abroad is typically from ten months to two years.2 By contrast, the market for foster care adoption is likely characterized by excess supply, as a large number of children in the foster care system are waiting to be adopted each year as we discuss below. 3. History of Child Adoption in the U.S., 1850-2000 To provide the historical background, we briefly describe the evolution of child adoption in the U.S. over the last 150 years.3 The development unfolded roughly in three stages. During the initial stage (circa 1850-1920), starting with the Massachusetts statute in 1851, an increasing number of states enacted adoption laws that established judicial supervision for the adoption of minors and 2 Country-specific information on adoption cost and time can be found at the U.S. State Department website: http://travel.state.gov/family/adoption/country/country_369.html. 3 For more details, see Moriguchi (2009). 5 enabled permanent legal transfer of parental rights from biological to adoptive parents upon court approval.4 Despite the efforts by charitable organizations to find adoptive homes for orphaned or abandoned children in large cities, adoptions had remained relatively few in number and informal in nature during the nineteenth century (Sokoloff (1993)). For example, in what is known as the “orphan train movement” in 1854-1904, about 100,000 homeless children in eastern cities were successfully placed on farm communities in the Midwest, but most of them were not formally adopted. The majority of homeless children in cities remained in the care of public or private institutions, where infants faced particularly high mortality rates due to infectious diseases and the lack of wet nurses. During the next stage (circa 1920-1960), the demand for adoption rose rapidly due partly to the improvements in infant formula, which enabled adoption of new-born babies and infants, and the growing perception that nurture, not nature, was the main determinant of child development. According to Albanesi and Olivetti (2007), the price of infant formula dropped dramatically in the 1930s and 1940s with the introduction of Similac, a leading brand that successfully replicated the nutritional contents of breast milk. By the 1950s, the demand for healthy infants in the U.S. began to exceed the supply (Lovelock (2000), p.912), and adoption agencies matched adoptable children with adoptive parents who shared similar socio-demographic characteristics. Most agencies limited their adoption practice to healthy infants of known background, and children with physical and mental disabilities were considered “unadoptable” well into the 1960s (Hansen (2006a)). During the third stage (circa 1960-2000), while the supply of adoptable infants within the U.S. declined rapidly, two additional sources of supply became available: foreign-born children and foster care children. First, responding to the large number of WWII orphans abroad, the federal government made the first endorsement of inter-country adoption in 1945 through special legislation, which was followed by similar one-time legislations (Lovelock (2000)). It was not until 1963, however, that Congress passed permanent legislation for inter-country adoption, establishing a special visa category for foreign children adopted by U.S. citizens. Since then, the federal government has gradually liberalized the criteria, increasing the age limit for children to be adoptable and allowing adoption by single parents (Weil (1984)). Second, as the number of children entering the foster care system increased substantially in the 1960s, there were concerns that these children were removed from their homes unnecessarily, and that once they entered the system, insufficient efforts were made to reunify them with their families or to find new adoptive 4 By contrast, in England, the first adoption statute was not enacted until 1926. 6 homes (Murray and Gesiriech (2004)). With the passage of the Adoption Assistance and Child Welfare Act of 1980, the federal government required all states to establish adoption subsidy programs to encourage the placement of foster care children. The act also established an adoption assistance program (“Title IV-E”) that provides states with federal matching funds to assist the adoption of special needs children. 4. Trends in Child Adoption in the U.S., 1944-2001 Our second goal is to survey available aggregate-level data and construct historical trends in U.S. adoption by type. As noted by NAIC (2004) and other adoption specialists, there is no single homogenous data source to estimate the number of children adopted in the U.S. each year. National estimates are made by Zarefsky (1946) and Maza (1984) based on state-level court data by the NCSS for the years 1944, 1951, 1955, 1957-1975. Flango and Flango (1995) combine special studies, court data, and vital records to estimate total adoption for the years 1987-1992. For more recent years, the National Council for Adoption (NCFA) conducted surveys in 1982, 1986, 1992, 1996, and 2002, and provide national estimates. Using similar but slightly different methods, NAIC (2004) also estimates total adoption for 2000 and 2001. Because these studies rely primarily on court records, the data include only formal adoption. The 2000 Census is the first census that asked the number of formally and informally adopted children residing in households, which provides the cumulative stock of adopted children in 2000 as opposed to the annual number of adoptions. No estimates for total adoption are available after 2002. As we discuss below, some micro surveys (e.g., NLSY, PSID, NSFG, and SIPP) do contain questions concerning adoption. Due to the low frequency of adoption, however, it is difficult to construct reliable time trends in adoption from these surveys. In Figure 1-a, we plot total number of adoption in the U.S. from 1944 to 2002 based on the estimates by Maza (1984), Flango and Flango (1995), NAIC (2004), and NCFA (1985, 1989, 1999, 2007). Note that these four separate series are not necessarily comparable due to differences in data and methods. Most important, there is a large discrepancy between the NAIC estimate for 2001 and the NCFA estimate for 2002, making it difficult to establish a recent trend. To resolve this issue, we construct upper bound estimates for the NAIC series and lower bound estimates for the NCFA series, respectively, using additional data (see “NCFA lower bound” and “NAIC upper bound” 7 series in Figure 1-a).5 The two sets of bounds overlap reasonably well, providing some assurance that the true values lie between these bounds. To control for changes in population and fertility over this period, in Figure 1-b, we present adoption rates, defined as the number of adoptions per 1,000 live births, from 1944 to 2002. Data limitations notwithstanding, the figure shows that adoption rates in the U.S. increased dramatically from less than 20 per 1,000 births in the early 1950s to over 45 per 1,000 births in 1968-1973. Adoption rates then declined sharply in the 1970s and possibly in the 1980s. As a result, adoption rates in 2002 (32.4 to 37.6 per 1,000 births) are still substantially lower than the historical peak reached in 1971 (47.5 per 1,000 births). To document the trends in inter-country adoption, Figure 2-a presents the number of foreign-born children adopted by U.S. citizens, 1945-2006. U.S. Immigration and Naturalization Service (renamed Department of Homeland Security in 2002) has reported annual data on inter-country adoption since 1962 with the establishment of special visa categories for immigrant-orphans.6 Before 1962, there were special one-time legislations in 1945, 1948, 1953, and 1957 that allowed a certain number of immigrant-orphans enter the U.S. and as such, these numbers do not represent annual inflows (Lovelock (2000); Weil (1984)). In Figure 2-b, we plot the inter-country adoption rate per 1,000 live births, 1962-2005. The figures show the rise and fall of inter-country adoption in three waves, first in the mid 1970s and second in the mid 1980s, seemingly uncorrelated with the trends in total adoption in Figure 1-b. Currently, we are in the midst of the third wave where the number of immigrant-orphans soared from 6,000 (or 1.6 per 1,000 births) in 1992 to over 20,000 (or 5.6 per 1,000 births) in 2004. In Figure 3, we decompose total adoption into related and unrelated adoption from 1951 to 2002 based on Maza (1984) and NCFA (1985, 1989, 1999, 2007). We also plot the estimates for unrelated adoption in 1976-85 by Bachrach et al. (1990) based on National Health Interview 5 The difference between the 2001 NAIC estimate and the 2002 NCFA estimate stems largely from the fact that, while the former assume that all inter-country adoptions are included in court data and vital records, the latter assumes that none are included (NCFA (2007), p.79, editor’s note). Foreign-born children adopted by U.S. citizens are included in these records only if they are adopted under U.S. state law. Those children who entered the U.S. under an IR4 visa are by federal law required to finalize their adoptions in a U.S. state court. Those who entered under an IR3 visa (whose adoption had been finalized in their birth countries) are not. Even so, government officials recommend IR3 children to be readopted in the U.S. to receive additional legal protection. Adoptive parents may incur nontrivial legal costs ($2,000 or higher) in doing so (CWIG (2004, 2006)). The number of IR3 and IR4 visa entrants are reported in USINS (1982-2002). No data are available as to how many IR3 children are readopted. We obtain lower bounds for the NCFA estimates by subtracting inter-country adoptions from their estimated total. Upper bounds for the NAIC estimates are obtained by adding IR3 adoption to their estimated total. 6 The 1962-2005 data do not include 2,911 children from Vietnam who were admitted to the U.S. in 1975 under a special refugee program (Weil (1984), p.289). 8 Survey (NHIS) data. Because the NHIS series match better with our NCFA lower bound series, we report the NCFA lower bound estimates in the figure. Assuming that all inter-country adoptions are unrelated adoption, we further divide unrelated adoption into domestic and inter-country unrelated adoption. Table 1 reports the share of related, unrelated, and inter-country adoptions in total adoption for selected years. The percentage of unrelated adoption in total was relatively stable at around 53\% in 1955-70, then dropped sharply from 51\% in 1970 to 37\% in 1975, and resurged recently from 36\% in 1982 to 58\% in 2002. The percentage of inter-country adoption in total increased steadily from 1.0\% in 1965 to 4.4\% in 1975, fluctuated between 4\% and 9\% in 1975-92, and then rose sharply from 5.1\% in 1992 to 13.9\% in 2002. This increase in inter-country adoption accounts for significant part of the recent surge in unrelated adoption, but not all of it. How large is stepchildren adoption? Although data are scarce, a large majority of related adoption appears to be stepparent adoption. In 1951, 39\% of all adoptions were by stepparents, 12\% were by relatives, and 48\% were unrelated; and in 1955, 35\% of all adoptions were stepparent adoptions, 10\% were by relatives, and 55\% were unrelated (NCSS (1951, 1955)). In 1992, among 26 reporting states, stepparent adoptions constituted an average of 42\% of all adoptions (Flango and Flango (1995)). To assess the share of foster care adoption in domestic unrelated adoption, in Figure 4, we divide domestic unrelated adoption by agency type (Maza (1984); NSFA (1985, 1989, 1999, 2007)). Public agency adoption refers to adoptions arranged by public child welfare agency, while non- public agency adoption includes private agency adoption and independent adoption. We also plot annual estimates for public agency adoption reported by the U.S. Children’s Bureau starting in 1990 (VCIS and AFCARS data in USBC (1990-2004)). The figure shows a significant rise in public agency adoption in the 1990s that coincides with the federal initiatives to encourage foster care adoption. In Table 2, we report the shares of public agency adoption and inter-country adoption in unrelated adoption for selected years.7 In 2002, 44\% of unrelated adoption is foster care adoption and 22\% are inter-country adoption. Domestic adoption through private agency or individual arrangements accounts for the remaining 34\% of unrelated adoption. Within foster care adoption, limited data suggest that special needs children (i.e., children over a certain age, of a sibling group, of minority race, or with disabilities) comprise a majority in recent years. From 1991-94, the average share of special needs children in public agency adoptions 7 No inter-country adoption is arranged through public agency. 9 ranged from 63\% to 82\% among the reporting states (USCB (1991-94)). In 2004, 47 states reported that 80 to 100\% of public agency adoptions were special needs adoptions (USCB (2004)). In summary, we find that, in the U.S., (1) the adoption rate in 2002, for both related and unrelated adoption, is still below its historical peak in 1968-73 despite a recent resurgence; (2) the share of unrelated adoption in total adoptions had declined sharply in the 1970s, but has increased over the last decade; and (3) this recent increase in unrelated adoption seems to be driven by the rise in inter-country and foster care adoptions. 5. Understanding Historical Trends in Adoption, 1951-2002 What explains the dramatic increase in the adoption rate for unrelated adoption in the 1960s and its equally dramatic decline in the 1970s? Why is the adoption rate today still substantially below the 1971 level? Why did inter-country adoption become a significant component of total adoptions only in the last decade despite its availability since the early 1960s? Can we attribute the rise in foster care adoption in the 1990s to the change in government policies? To what extent, are inter- country adoption and foster care adoption a substitute for domestic private agency adoption? To better understand the historical trends, we divide the U.S. adoption market into three segments, (1) domestic private agency adoption, (2) inter-country adoption, and (3) foster care adoption, and explore possible demand- and supply-side explanations using historical data. 5.1 Market for Domestic Private Agency Adoption We first consider the demand-side factors that affect the market for domestic private agency adoption that deals primarily with unrelated adoption of healthy infants. [Revise: also a substantial number of unrelated adoption of infants by public agency and independent arrangements.] A dramatic rise in women’s educational attainment and labor force participation since the 1950s has been associated with delayed marriage and childbearing (Goldin and Katz (2002); Caucutt et al. (2002); Olivetti (2006)). Figure 5 shows that the labor force participation rates for women aged 20-44 had increased sharply from the early 1960s to the late 1980s. Figure 6 documents a concurrent increase in the age of women at first marriage and at first birth. The dramatic rise in women’s educational and occupational attainment implies that interrupting work to bear a child incurs a higher opportunity cost. Indeed, recent studies have found substantial wage premium on delayed childbearing, especially for college educated women and women in highly skilled 10 professions (Ellwood et al. (2004); Miller (2006); Buckles (2006a)). 8 Delayed motherhood, however, is in turn associated with higher risk of facing infertility before achieving a desired number of children.9 As more women seek both career and family in recent decades (Goldin (2004)), we expect a potentially large increase in the demand for adoptable infants as a substitute for childbearing. In fact, preceding studies have consistently found that women’s inability or difficulty in bearing a child is positively related with their likelihood of adoption.10 At the same time, recent progress in infertility treatment has improved the probability of having biological children for women with fertility problems. Two major advancements are the 1967 FDA approval of fertility drugs, which induce ovulation, and the 1981 introduction of in vitro fertilization (IVF), the most common form of assisted reproductive technology (ART) today.11 As the usage of fertility drugs correlates with the incidents of multiple births, the diffusion of fertility drugs can be inferred, albeit imperfectly, from the changes in multiple birth rates. Figure 7 plots the ratio of triplet and higher-order multiple births per 100,000 live births in the U.S. from 1971 to 2002. The ratio was nearly constant in the 1970s despite the earlier approval of fertility drugs, and then dramatically increased from the early 1980s to the late 1990s.12 Figure 8 presents the number of ART cycles performed and the numbers of resulting live births and deliveries in the U.S. from 1985 to 2003 (note that one delivery may produce multiple births). The figure also plots the success rate, measured by the percentage of ART cycles resulting in live deliveries. The number of ART deliveries has increased dramatically from 3,951 in 1990 to 25,228 in 2000, and the success rate doubled from 13\% to 25\% during the same period. Improvements in ART led to both the reduction in the monetary cost per delivery and the decline in the risk of multiple births over the last decade 8 Miller (2006) finds that an additional year of fertility delay is associated with a 3\% increase in hourly wage rates and a 10\% increase in lifetime earnings for women. 9 For example, the probability of conceiving and delivering a healthy baby for non-contracepting women declines by half from age 25 to age 35 (Van Noord-Zaadstra et al. (1991)). 10 See Bonham (1977), Bachrach (1983, 1986), Bachrach et al. (1990), Bachrach et al (1991), Chandra et al. (1999), and Hollingsworth (2000). Chandra et al. (1999), however, suggest that this relationship may be weakening over time. 11 ART refers to procedures that involve retrieving eggs from ovaries, combining them with sperms in the laboratory, and transfer them into a woman’s uterus or fallopian tube. Artificial insemination, which is not part of ART, has been used to treat infertility since the pre-WWII period with relatively minor technological improvements. 12 The figure overstates the diffusion of fertility drugs because multiple births increase also with maternal ages and the use of ART (Martin and Park (1999)). 11 (Toner (2002)). Nevertheless, estimated costs of IVF per delivery in 2006 remain very high, ranging from $30,000 to $ 60,000 in 2006.13 To what extent, are advanced infertility treatment and adoption substitutes? The ratio of the number of women who delivered their biological children with ART to the number of women who adopted unrelated children domestically has increased from 15\% in 1992, 34\% in 1996, and to 60\% in 2002 (based on the NSFA lower bound estimates for domestic unrelated adoption). In other words, ART likely had a sizable impact on the demand for domestic infants in recent years. To summarize, the continuing trend in delayed childbearing has likely increased the demand for adoptable infants from the 1960s to the 1990s. Starting in the 1980s, however, advancement in ART likely reduced the adoption demand particularly from those individuals with higher income or stronger preference for biological children. If the adoption market for …